--- title: Webinar | Building More Accurate Spatial Models with Google PDFM Embeddings and CARTO description: Learn how to improve the accuracy of your predictive models by integrating Google’s Population Dynamics Foundation Model (PDFM) embeddings in CARTO. Explore how to enrich spatial analysis with behavioral and environmental patterns. image: https://go.carto.com/hubfs/Webinar_Building%20More%20Accurate%20Spatial%20Models%20with%20Google-3.png --- - [Log in](https://carto.com/signin/) - [Talk to us](https://carto.com/request-live-demo/) RECORDED WEBINAR ![CARTO white (2)-1](https://go.carto.com/hubfs/CARTO%20white%20(2)-1.svg) Building More Accurate Spatial Models with Google PDFM Embeddings and CARTO ![](https://go.carto.com/hubfs/Miguel%20Alvarez.jpeg) ###### Miguel Alvarez Lead Data Scientist at CARTO ![](https://go.carto.com/hubfs/Gautam%20Prasad.jpeg) ###### Gautam Prasad Staff Software Engineer at Google Research Join us for a technical session focused on enhancing the precision of your predictive models. While traditional demographic data provides a solid foundation for spatial analysis, it often lacks the nuance to capture real-world population dynamics, leading to gaps in model accuracy.   This webinar introduces a powerful solution: the integration of Google's Population Dynamics Foundation Model (PDFM) embeddings directly with CARTO.   Learn how these high-dimensional vector representations encapsulate complex behavioral and environmental patterns, providing a richer, more dynamic layer of geographic insight.   We will demonstrate through a live demo how you can seamlessly incorporate these pre-trained features into CARTO Workflows to build more accurate and robust predictive models for use cases like sales forecasting and resource allocation.  ![](https://www.facebook.com/tr?id=273642683065267&ev=PageView%20&noscript=1)